How could I best determine the value of a suitable capacitor (if this is the best way) to smooth the results as I am getting:
20.8582248
20.8597068
20.8745136
20.8471202
20.8708133
20.8374958
20.8619270
20.8256511
20.8597068
20.8552646
20.9181957
20.8367557
at 200ms intervals.
Would the capacitor be best in parallel over the contacts or to ground?
Many thanks ps this is based on the LTC2400 circuit I am using just for now.
Accidentally ordered ina114!
I will update once it arrives.
I used an ADS1115 and managed to get a resolution of 0.02% with was ok but I need 0.01 and this was running it at its maximum 16x pga and single ended so:
16bit must be 65535 and at a 0.0078125mV resolution my range (0-40mv) spans only a quarter of the range???
OK, so I finally get an INA112 in-amp.
I look at the datasheet and see that a NC of the gain resistor results in a 5x gain so nice and easy, I will leave this out for the purposes of a test.
I connect the oxygen sensor to pins 2 and 3, 4 to ground, 5 to ground and 7 to 5v. I connect pin 6 to a multimeter.
I get a relatively stable 4.7 volts as if the op-amp is going out of range and hitting its limit.
What am I doing wrong?
Thanks
Hmmm, you might try connecting pin 2 (or 3 ??, but not both obviously) to ground as well. IIRC I had problem in the sim when the reference voltage was different from the pin 2 (or 3 ??) voltage. The both needed to be biased off of ground for the INA114, 'cuz it’s not a “rail to rail” op-amp.
Also try reversing the connections to the sensor so it outputs a 0 to + voltage (as known by the op-amp) and not a 0 to - voltage.
EDIT : Here’s how I had it connected in the sim (powered from a 9v battery) and it made a nice linear plot from from 1.5 V to 5.0 V w/the sensor output going 0 - 55 mV. So the -in can be biased different from the reference voltage but it did have to have some tie to something between Vcc and ground (perhaps just ground). Leaving that connection open stuffed Vout into the (bottom) rail.
(note the 5v labels are really 8.5v)